A. Berke Gökmen

Generative Modeling & 3D Perception

I'm a researcher at INSAIT where I prototype controllable diffusion systems for human- and scene-centric understanding.

Previously at ETH Zürich's Computer Vision Lab and Bilkent's Deep Learning Lab, I build collaborative pipelines that push generative ideas into deployable products.

Publications

NeurIPS 2025 RoPECraft: Training-Free Motion Transfer with Trajectory Guided Optimization on DiTs

Ahmet Berke Gökmen*, Yiğit Ekin*, Bahri Batuhan Bilecen*, Ayşegül Dündar (* joint first authors)

Paper Project Code
NeurIPS 2024 Dual Encoder GAN Inversion for High-fidelity 3D Head Reconstruction from Single Images

Ahmet Berke Gökmen*, Bahri Batuhan Bilecen*, Ayşegül Dündar (* joint first authors)

Paper Project Code
ICCV 2025 Identity Preserving 3D Head Stylization with Multiview Score Distillation

Bahri Batuhan Bilecen, Ahmet Berke Gökmen, Furkan Güzelant, Ayşegül Dündar

Paper Project Code
arXiv 2025 Counterfactual Disease Removal and Generation in Chest X-Rays Using Diffusion Models

Ahmet Berke Gökmen, Ender Konukoğlu

Paper Project Code

Experience

Roles

INSAIT — Researcher · Aug 2025 – Present · Sofia
ETH Zürich CVL — Research Intern · Jul 2024 – Nov 2024 · Zürich
Bilkent DLR — Undergraduate Researcher · Jan 2024 – Aug 2025 · Ankara
Getir — Software Engineer · Sep 2022 – Jan 2024 · Istanbul
CMMRS — Visiting Student · Aug 2025 · Saarbrücken

Education

Bilkent University — B.E. Computer Science · 2020 – 2025
EPFL — Exchange, Computer Science · 2023 – 2024
Çanakkale Özel Fen Lisesi — Math & Science Diploma